276 



quantity is divided equally between the three pipes in each trench. 

 If this method of fumigation is carefully carried out during the west 

 monsoon, the complete destruction of the gramang ant is assured, 

 and if neighbouring estates are also treated there will be no re-infesta- 

 tion. The carbon bisulphide may also be poured into holes made 

 by thrusting a stick into the leaves, but the bamboo pipes increase 

 the efficacy and economy of this method. 



Burying the infested leaves in trenches was only successful when 

 the trenches were 18 inches deep. At this depth the result was 

 tolerably good, but still inferior to that with carbon bisulphide, and, 

 as this method aims at suffocating the insects, it must be practised 

 in the rainy season ; the experiments were conducted during the 

 rains and in clay soil. Gasoline and another by-product of petroleum 

 refining, aerogine, proved much less efficient than carbon bisulphide. 

 This also applies to potassium cyanide, even in a solution of 2 per 

 cent, strength, and to formalin. Banding experiments were made 

 during the rains and a dry period lasting a few weeks. Under those 

 conditions a material obtained from Amsterdam retained its 

 adhesiveness for more than two months. Banded coffee bushes were 

 entirely free from ants unless they were prevented from going else- 

 where, in which case some of the workers crossed the banding. Up 

 to the time of writing, no injury had been caused by applying the 

 adhesive direct on the trunk instead of on a band of paper. Banding 

 is suggested as a likely means of dealing with ants which are not so 

 susceptible as P. longipes to control by means of trap-trenches and 

 it is recommended against Mynnicaria brunnea, the nests of which 

 are deep under ground. 



Some supplementary notes are given on Coccus viridis, chiefly 

 dealing with the points in which the author compares his experience 

 with information given by Keuchenius. The latter has stated that 

 the developmental period (by which the present author means the 

 time up to the production of the first progeny) occupies about 45 

 days, whereas observations have showni it to last at least 65 days, 

 thus proving that considerable individual variations occur. Again, 

 the green scale increases in four months as from 1 to 400 and this 

 increase, though great, is considerably lower than that stated by 

 Keuchenius. The latter has pointed out that this scale attains a 

 larger size on coffee than on Gardenia fiorida. It may be added that 

 on coffee alone does this insect really thrive. Keuchenius has also 

 stated that new food-plants are gradually being attacked, but the 

 view here taken is that the increase observed is due to the greater 

 attention now given to the subject. The following plants, not recorded 

 by Keuchenius, are general and important hosts of this scale in central 

 Java : — Citrus aurantium (orange), Justicia betonica, Plumiera acuti- 

 folia and Psidium guajava (guava). 



Indigenous insect enemies of the green scale include : — The 

 Coccinellids, Chilocorus melano'phthalmus, Muls., Orcus janthinus, 

 Muls., and the Chalcids, Cocco'phagus bogoriensis, Koningsb., 

 Myiocnema comperei, Ashm., Aneristus ceroplastae, How., Coccophagus 

 sp., Cheiloneuromyia javensis, Gir., Cristatithorax Jatiscapvs, Gir. 

 These species are briefly described and accounts of their life-histories 

 are given. 



